( Hordeum Vulgare Var. Himalaya 292) Lowers Plasma Cholesterol And

Downloaded from

https://www.cambridge.org/core

British Journal of Nutrition (2004), 92, 607–615 DOI: 10.1079/BJN20041248 q Commonwealth Scientiﬁc and Industrial Research Organisation 2004

A novel high-amylose barley cultivar (Hordeum vulgare var. Himalaya

. IP address: 292) lowers plasma cholesterol and alters indices of large-bowel fermentation in pigs

170.106.40.139

Anthony R. Bird*, Michelle Jackson, Roger A. King, Debra A. Davies, Sylvia Usher and David L. Topping , on

27 Sep 2021 at 03:13:57 CSIRO Health Sciences and Nutrition, PO Box 10041, Kintore Avenue, Adelaide BC 5000, Australia

(Received 3 February 2004 – Revised 28 June 2004 – Accepted 30 June 2004)

Hordeum vulgare var. Himalaya 292 is a new barley cultivar with altered starch synthesis and less total starch but more amylose, resistant , subject to the Cambridge Core terms of use, available at starch (RS) and total and soluble NSP including b-glucan. To determine its nutritional potential, young pigs were fed diets containing stabilised wholegrain flours from either Himalaya 292, Namoi (a commercial barley), wheat bran or oat bran at equivalent dietary NSP concentrations for 21 d. Serum total cholesterol was significantly lowered by the Himalaya 292 diet relative to wheat bran, indicating that Himalaya 292 retained its hypocholesterolaemic potential. In all groups SCFA concentrations were highest in the proximal colon and decreased towards the rectum. Digesta pH was lowest in the proximal colon and highest in the distal colon. Large-bowel and faecal pH were significantly lower in the pigs fed the barley and oat diets, indicating greater bacterial fermentation. Caecal and proximal colonic pH was lowest and SCFA pools highest in the pigs fed Himalaya 292. Total and individual SCFA were lowest in the mid- and distal colon of the pigs fed Himalaya 292 or oat bran. These data suggest the presence of more RS in Himalaya 292 and suggest that its fermentation was rapid relative to transit. Differences in faecal and large-bowel anaerobic, aerobic, coliform and lactic acid bacteria were relatively small, indicating a lack of a specific prebiotic action. These data support the potential of this novel barley cultivar to improve health through plasma cholesterol reduction and increased large-bowel SCFA production.

Bacteria: Dietary ﬁbre: Plasma cholesterol: Starch: Barley

https://www.cambridge.org/core/terms

Plant foods and their carbohydrate components have well- function, such as the stimulation of fluid and electrolyte established roles in the management and prevention of absorption and the modulation of muscular contraction non-infectious chronic illnesses. Wholegrain cereal con- and visceral perfusion (Topping & Clifton, 2001). It appears sumption has been linked to a lowered risk of certain cancers that the intakes and fermentability of NSP are insufficient to and of CHD (Marquart et al. 2002; Truswell, 2002). These support the bacterial colonisation of the normal human large cereals are high in starch and dietary fibre. The latter bowel. This is the, so-called, ‘carbohydrate gap’, which is comprises largely NSP (Spiller, 2001), and NSP-rich foods filled by other substrates including RS (Stephen et al. 1983). are well-established laxating agents (Borum, 2001). Soluble RS is a modest faecal bulking agent and some studies

NSP, such as b-glucan and those of psyllium, have been suggest that a combination of NSP and RS is especially . shown to lower serum cholesterol (Brown et al. 1999; useful for the promotion of large-bowel health (Govers https://doi.org/10.1079/BJN20041248 Anderson et al. 2000; Jenkins et al. 2002), an established et al. 1999). The modification of foods to enhance the risk factor for early-onset CHD. NSP resist human intestinal content and action of RS and NSP is a recognised means digestive enzymes, which helps to explain their effective- of effecting dietary change at the population level to ness in increasing faecal bulk, an important component of improve public health. A novel cultivar of a hull-less their laxative action (Topping & Clifton, 2001). Resistant barley (Hordeum vulgare var. Himalaya 292) with a starch (RS) is that fraction of starch and the products of its specific defect in starch synthesis has been identified digestion that are not absorbed in the small intestine of recently and grown in quantity (Morell et al. 2003). This healthy human individuals and enter the large bowel (Asp, modification occurs in starch synthase IIa (EC 2.4.1.21) 1992). In the latter viscus both RS and a variable fraction and results in a shrunken grain containing as little as of NSP are fermented by the microflora (Topping & Clifton, 18 % starch, of which approximately 70 % is present as 2001). SCFA are major products of this fermentation and amylose, compared with approximately 50 and 25 % they promote several important aspects of large-bowel respectively for its Himalaya parent (Morell et al. 2003).

Abbreviations: RS, resistant starch; TAG, triacylglycerol. * Corresponding author: Dr A. R. Bird, fax þ61 8 8303 8899, email [email protected] Downloaded from

https://www.cambridge.org/core

608 A. R. Bird et al.

Himalaya 292 grain contains approximately 25 % total (designated week 1), all pigs were maintained on the NSP, and 10 % b-glucan (Topping et al. 2003). These com- wheat-bran control diet (Table 1), after which they were positional changes suggested that Himalaya 292 would be blocked on live weight and assigned randomly within high in RS (Morell et al. 2003), and this was conﬁrmed in a blocks to one of four dietary treatment groups, each of feeding trial in rats, which showed substantially greater seven animals for a period of 3 weeks (designated weeks . IP address: caecal starch in rats fed Himalaya 292 compared with 2, 3 and 4). The experimental diets were formulated to wheat bran, oat bran and a commercial barley cultivar supply equal quantities of fat, protein and carbohydrate

(Bird et al. 2004). The rat has limitations as an experimen- in the proportions 8, 20 and 60 % respectively. The fibre 170.106.40.139 tal model for human plasma lipid metabolism and large- content (as NSP) was set at 75 g/kg diet and was supplied bowel carbohydrate digestion and fermentation (Roe et al. as wheat bran (wheat-bran diet; control), Himalaya 292 1996). In the present study, therefore, we sought to explore wholemeal flour, standard hull-less barley wholemeal the nutritional properties of Himalaya 292 in a more rele- flour (Namoi) or a commercial oat-bran product (Oat , on vant species, the pig. Specifically we sought to determine Golde; oat-bran diet). The proximate composition of 27 Sep 2021 at 03:13:57 whether the processed grain lowered plasma cholesterol Himalaya 292 and the other cereal ingredients has been relative to wheat bran. We also assessed whether there reported previously (Bird et al. 2004). All diets were were any changes in faecal and large-bowel SCFA and balanced for their macronutrient content by changing the microflora. proportions of the other ingredients. The starch component of the control diet was provided by a pre-gelatinised low-

amylose starch and this same product was used where , subject to the Cambridge Core terms of use, available at Materials and methods necessary to balance the starch level in the other diets. Care of animals Pigs had free access to fresh water, and diets were meal- fed at 70 g/kg live weight0·75. They were weighed weekly The study was carried out under contract at the Pig Metab- before the morning feeding and food intake adjusted olism Unit of the Victorian Institute of Animal Science at accordingly. Dry rations were mixed with water in a 2:1 Werribee, Victoria, Australia. The study was approved by ratio (w/w) and offered twice daily as a morning and eve- the relevant institutional animal experimentation commit- ning meal. Cooked minced meat was weighed separately tees and conformed to published guidelines (National and then mixed with the wet diet. Wheat bran with a par- Health and Medical Research Council, 1997). ticle size of 2–3 mm and suitable for human consumption was supplied by Bartlett Grains (Port Adelaide, South Aus- tralia, Australia). Pre-gelatinised low-amylose starch Diets, feeding and experimental design (B012XF) was obtained from Penford Australia Pty Ltd Twenty-eight male crossbred pigs (about 9 weeks of age) (Lane Cove, NSW, Australia); Namoi and Himalaya 292 with initial live weights averaging 24 kg were used in a were obtained from CSIRO Plant Industry (Black Moun- randomised complete block design experiment. They tain, ACT, Australia) and were heat stabilised batch-wise https://www.cambridge.org/core/terms were housed in individual galvanised steel pens with con- using moist heat at 100–1208C, with steam delivered at crete ﬂoors in a pig grower shed and were allowed 7 d to 110 kPa for 12 min, followed by vacuum at 280 kPa for adapt to their surroundings and feeding regimen before 20–30 min. Oat Golde was supplied by The Uncle starting the experiment. During this adjustment phase Toby’s Company (Wahgunyah, Victoria, Australia).

Table 1. Ingredient composition of control and experimental diets (g/kg)

Diet

. Ingredient Wheat bran Namoi Himalaya 292 Oat bran https://doi.org/10.1079/BJN20041248

Wheat bran 162 Namoi 608 Himalaya 292 256 Oat Golde 287 Cooked minced beef 229 139 194 186 Skimmed milk powder 159 97 135 129 Starch* 363 62 341 329 CaHPO4·2H2O 4·4 4·4 4·4 4·4 Sucrose 44 44 39 44 Safﬂower-seed oil 34 40 26 15 NaCl 2·2 2·2 2·2 2·2 Vitamins and minerals† 1·5 1·5 1·5 1·5 Chromic oxide 1·0 1·0 1·0 1·0

* Pre-gelatinised low-amylose starch (B012XF; Penford Australia Pty Ltd, Lane Cove, NSW, Australia). † Pharmamix P169 (Propharma Australia Pty Ltd, Dandenong, Victoria, Australia), which contained (per kg mix): retinyl acetate, 1·5 g; cholecalciferol, 25 mg; a-tocopherol, 20 g; riboﬂavin, 2 g; cyanocobalamin, 7·5 mg; calcium pantothenate, 5·6 g; biotin, 50 mg; nicotinamide, 10 g; menadione, 1 g; FeSO4·7H2O, 50 g; MnO2, 10 g; ZnO, 50 g; CuSO4·7H2O, 5 g; CoSO4, 0·25 g; KI, 0·5 g; Na2SeO4, 0·1 g; antioxidant, 31 g (Oxicap E2; Novus Nutrition, Melbourne, Australia). Downloaded from

https://www.cambridge.org/core

Effect of novel barley cultivar in pigs 609

Sampling and analytical procedures Statistical methods Statistical evaluations of the data were performed by two- Faecal samples were collected during each week of the way ANOVA using the general linear model procedure of study. Freshly voided stools were frozen for subsequent

SAS (version 8.02; Statistical Analytical Systems Institute . IP address: analyses of SCFA, moisture and pH. Faecal samples Inc., Cary, NC, USA). The model included diet and weight for bacteriological enumeration were diluted 8-fold with block as sources of variation. When significant values were aqueous storage medium (buffered peptone (20 g/l), detected for diet (F value P,0·05), differences between lab-lemco powder (20 g/l) and 20 % (v/v) glycerol) and 170.106.40.139 individual means were then analysed by the protected frozen (2208C). For analyses, suspensions were thawed difference procedure. Data are shown as least squares at 48C and further serially diluted in pre-reduced means and pooled standard errors of the least squares buffered peptone (buffered peptone (20 g/l; Oxoid, means (SEM). Where there is a missing value this is Australia Pty Ltd, West Heidelberg, Victoria, Australia), , on noted in the Tables and the SEM is weighted to allow for 0.5 % cysteine HCl and 0.1 % Tween 80). Portions 27 Sep 2021 at 03:13:57 this change. Values in any Table column with unlike super- (100 ml) of the appropriate dilutions were inoculated in script letters were significantly different (P,0·05). For duplicate onto plates containing Bifidus blood (Pachenari counts of bacteria, log transformation of the data was et al. 2001), Rogosa (Oxoid), Columbia blood (Oxoid) 10 performed before the statistical determination of treatment and Chromogenic Escherichia coli–coliform (Oxoid) effects. In one instance (faecal bifidobacteria numbers), media for the selective enumeration of Bifidobacteria, analysis showed a significant difference between groups Lactobacilli, total anaerobes, E. coli, total coliforms, , subject to the Cambridge Core terms of use, available at at week 1. In this instance, analysis of the values was con- and total aerobes respectively. Plates were incubated at ducted with initial values as a covariate. 378C under anaerobic conditions using Anaerogen Compact (Oxoid) for a period of 5 d for Bifidus blood and Columbia blood plates and 3 d for Rogosa plates. Chromogenic plates were incubated at 378C under Results aerobic conditions for 24 h. Colonies characteristic of Feed intake and body-weight gain each bacterial group were then counted visually and the concentration calculated as colony-forming units/g All diets were accepted and all food was eaten promptly on wet weight. presentation. The amount of food offered was set as a con- At the completion of the treatment period, and about stant fraction of body weight and there were no significant 16 h after the pigs had been fed their final meal the pre- differences in body weight at the beginning of the experi- vious evening, they were weighed and anaesthetised with ment or at any point throughout the trial. Mean initial halothane. The abdominal cavity was then opened and body weight for all groups combined was 24·5 (pooled blood was withdrawn from the abdominal aorta and SEM 0·2) kg and mean final body weight was 32·0 serum prepared and stored for the later analysis of (pooled SEM 0·5) kg. https://www.cambridge.org/core/terms lipids. Pigs were then euthanased by an intracardial injection of an overdose of barbiturate. The entire gastrointestinal tract was tied off immediately, excised and the Serum lipids large bowel and small intestine separated and their The mean concentration of serum total cholesterol in the length measured. The caecum and colon were ligated pigs fed Himalaya 292 was significantly lower than in the and the latter divided into three segments of equal pigs fed wheat bran. The mean concentration in the pigs length while the small intestine was divided into ten seg- fed oat bran was similar to that in the pigs fed Himalaya ments of equal length. Digesta from each of the small- 292, but this was not significantly different from wheat and large-intestinal segments was extruded and weighed. bran; the mean concentration in the pigs fed Namoi was

. The colon and caecum, empty of contents, were blotted intermediate to those of wheat bran and Himalaya 292 https://doi.org/10.1079/BJN20041248 dry and weighed. Samples of digesta from the proximal (Table 2). HDL concentrations did not differ between colon were collected using aseptic techniques and groups so this difference was attributable entirely to a low- diluted 1:9 (w/v) in storage medium and bacteria enum- ering of VLDL- þ intermediate-density lipoprotein- þ erated as described for faeces. Additional colonic digesta LDL-cholesterol (calculated by difference). Serum TAG samples were stored at 2208C for later determination of were unaffected by diet (Table 2). DM and pH using standard procedures. Faecal and digesta SCFA were isolated by low-temperature vacuum microdistillation and quantified by GLC (Bird Faecal pH and short-chain fatty acids et al. 2000b). Briefly, frozen faecal and digesta samples were thawed at 48C and a sub-sample diluted with hep- There were no significant differences in faecal pH between tanoic acid as internal standard, centrifuged, acidified groups at the start of the feeding period (week 1; data not and distilled under vacuum and SCFA in the distillates shown). Values did not change significantly throughout the measured by GLC. Total cholesterol, HDL-cholesterol, study in the pigs fed wheat bran but pH was significantly and triacylglycerol (TAG) were measured in serum (and equally) lower in the other groups at week 4 using commercial kits (Roche Diagnostics, Mannheim, (Table 3). There were no significant differences between Germany) and a Cobas-Bio centrifugal analyser (Roche, groups for faecal total and individual SCFA concentrations Basel, Switzerland). at the start of the study and at week 4 (Table 3). Downloaded from

https://www.cambridge.org/core

610 A. R. Bird et al.

Table 2. Serum lipids in pigs after consumption of experimental and wheat-bran (control) diets for 3 weeks (seven pigs per diet)* (Mean values and pooled standard errors of the mean)

Lipid concentration (mmol/l) . IP address:

Diet Total cholesterol VLDL- þ LDL- þ IDL-cholesterol HDL-cholesterol TAG

a a

Wheat bran 3·06 1·79 1·27 0·79 170.106.40.139 Namoi 2·87a,b 1·59a,b 1·28 0·90 Himalaya 292 2·58b 1·39b 1·19 0·79 Oat bran† 2·62a,b 1·42a,b 1·20 0·89 Pooled SEM 0·16 0·14 0·06 0·09

, on IDL, intermediate-density lipoprotein; TAG, triacylglycerol. a,b Mean values within a column with unlike superscript letters were signiﬁcantly different (P,0·05). 27 Sep 2021 at 03:13:57 * For details of diets and procedures, see Table 1 and p. 608. † Mean values for six pigs.

Table 3. pH and short-chain fatty acid concentrations in faeces of fed wheat bran (Table 4). In contrast to all other major bac- pigs after consumption of experimental or wheat-bran (control) diets terial groups, Bifidobacteria numbers were significantly for 3 weeks (seven pigs per diet)* lower at week 1 in the pigs assigned to be fed Namoi com- , subject to the Cambridge Core terms of use, available at (Mean values and pooled standard errors of the mean) pared with those assigned to be fed wheat bran (4·95 v. 7·10 colony-forming units/g, respectively; P,0·01) and Short-chain fatty acid at week 4 compared with both wheat bran and Himalaya concentration (mmol/kg) 292 (Table 4). This latter difference remained significant Diet pH Acetate Propionate Butyrate Total even when the data were analysed with the week 1 values as covariate. Wheat bran 6·84a 43·1 17·7 8·4 69·2 Namoi 6·54b 52·0 23·2 11·2 86·3 Himalaya 292 6·49b 43·8 15·7 8·0 67·6 Oat bran 6·43b 39·7 16·6 8·9 65·2 Intestinal length and weight, and digesta weight Pooled SEM 0·05 4·9 3·5 1·5 9·3 Neither the length nor the weight of small or large intestine a,b Mean values within a column with unlike superscript letters were signifi- free of content was affected by diet (data not shown). The cantly different (P,0·05). wet weight of small-intestinal digesta was lowest in the * For details of diets and procedures, see Table 1 and p. 608. pigs fed the Namoi diet and significantly higher in those fed oat bran with respective mean values of 340 and 448

Faecal bacteria https://www.cambridge.org/core/terms (pooled SEM 35) g (P,0·05). Values were 366 and 404 g Faecal anaerobe numbers did not differ between groups at for the wheat-bran and Himalaya 292 groups, respectively. week 1 (data not shown) but were significantly lower in the These did not differ from each other or the other treat- pigs fed wheat bran than in those fed Himalaya 292 and oat ments. There were significant effects of diet on large- bran at week 4 (Table 4). Faecal aerobe numbers were not bowel digesta with significantly more in the pigs fed significantly different between groups either at week 1 or at wheat bran than in all other groups (Table 5). There was week 4. There were no significant differences in the num- also a significant interaction between diet and sampling bers of total coliforms and E. coli between groups at any site, with more digesta in the mid- and distal colon of sampling time (data not shown). In contrast there were sig- the pigs fed wheat bran than in those fed Himalaya 292 nificant effects of diet on faecal lactic acid bacteria. Faecal or oat bran. Digesta water content was unaffected by diet

. numbers of Lactobacilli were similar between groups at the and fell signiﬁcantly (P,0·001) in all groups from https://doi.org/10.1079/BJN20041248 initial sampling time but at week 4 numbers were signiﬁ- approximately 86 % in the caecum to 75 % in the distal cantly lower in the pigs fed Namoi compared with those colon (data not shown).

Table 4. Bacteria numbers (log10 colony-forming units/g) in faeces of pigs after consumption of experimental and wheat-bran (control) diets for 3 weeks (seven pigs per diet)* (Mean values and pooled standard errors of the mean)

Diet Anaerobes Aerobes Lactobacilli Biﬁdobacteria

Wheat bran 9·13a 7·94 8·20a 7·47a Namoi 9·23a,b 7·66 7·35b 6·48b Himalaya 292 9·44b 8·00 8·16a,b 7·51a Oat bran 9·44b 8·05 7·49a,b 7·33a,b Pooled SEM 0·07 0·14 0·27 0·29

a,b Mean values within a column with unlike superscript letters were signiﬁcantly different (P,0·05). * For details of diets and procedures, see Table 1 and p. 608. Downloaded from

https://www.cambridge.org/core

Effect of novel barley cultivar in pigs 611

Table 5. Large-bowel digesta weight (g) of pigs after consumption Concentration is a relatively limited indicator of large- of experimental or wheat-bran (control) diets for 3 weeks (seven bowel SCFA status, as it gives no indication of their total pigs per diet)* availability. Pool (i.e. concentration £ digesta volume) is a (Mean values and pooled standard errors of the mean) much better indicator. As expected from the concentration data, there were signiﬁcant independent and interactive . IP address: Colon effects of diet and sampling site on large-bowel SCFA Diet Total Caecum Proximal Mid Distal pools. Individual comparisons revealed that proximal colo-

nic digesta total SCFA pool was greatest in the pigs fed 170.106.40.139 Wheat bran 772a 128 307 203a 132a Namoi 615b 132 277† 158a,b 101a,b Himalaya 292 (Table 8) but this difference was signiﬁcant Himalaya 292 620b 142 299 112b 68b,c only compared with oat bran. In the mid- and distal colon, Oat bran 511b 89 244 125b 54c SCFA pools were similar in the pigs fed wheat bran and

Pooled SEM 50 23 26 18 12 Namoi, and significantly greater in the pigs fed Namoi , on a,b,c than in those fed Himalaya 292. In the distal colon, pools 27 Sep 2021 at 03:13:57 Mean values within a column with unlike superscript letters were significantly different (P,0·05). in both wheat-bran and Namoi groups were equally and sig- * For details of diets and procedures, see Table 1 and p. 608. nificantly greater than in the other two groups. As expected, † Mean value for six pigs. differences in acetate pools followed very similar patterns to those of total SCFA. As with acetate, the pools of propionate Large-intestinal pH and short-chain fatty acid were higher in the proximal large bowel of the pigs fed concentrations and pools Himalaya 292 (and significantly greater than in those fed , subject to the Cambridge Core terms of use, available at oat bran) (Table 8). In contrast, in the distal colon there Concentrations of total caecal SCFA were highest, and pH were lower values in the pigs fed Himalaya 292 or oat values lowest, in the pigs fed Himalaya 292 and were sig- bran compared with Namoi while in the mid-colon propio- nificantly different from those of the pigs fed wheat bran nate pools were smaller in the pigs fed Himalaya 292 com- (Tables 6 and 7). Similar differences in SCFA concen- pared with Namoi. Broadly similar differences were trations were seen in the proximal colon. There were no obtained with butyrate (Table 8). significant differences in the mid-colon but in the distal colon, total concentrations of SCFA were higher in the Namoi group than in the oat-bran or Himalaya 292 Proximal colonic bacteria groups. Acetate had a regional concentration profile very Total anaerobe numbers were highest in the pigs fed Hima- similar to that of total SCFA. Thus, acetate was highest laya 292 but the difference was significant only against in the caecal digesta of the pigs fed Himalaya 292 and wheat bran (Table 9). Aerobe numbers were significantly in the proximal colonic contents of those fed Namoi lower in the animals fed Namoi than in all other groups. (Table 7). Caecal and proximal colonic propionate concen- There were no treatment effects on numbers of Lactobacilli trations were significantly higher in the pigs fed Himalaya or Bifidobacteria. Digesta total coliform and E. coli num- https://www.cambridge.org/core/terms 292 than in the pigs fed wheat bran (Table 7). In the mid- bers were lowest in the pigs fed Namoi than in all of the colon, the difference was significant only between wheat other groups and these differences were significant against bran and Namoi. In the distal colon, concentrations were all other treatments for coliforms and against Himalaya higher in the pigs fed Namoi than in all other groups, 292 and oat bran for E. coli (Table 9). although this was statistically significant only for Himalaya 292 and oat bran. Concentrations of butyrate were highest in the caecum of the pigs fed Himalaya 292 but the differ- Discussion ence was statically significant only against wheat bran The genetic change in Himalaya 292 leads to substantial (Table 7). There were no significant differences in concen- alterations in starch content and composition (Morell tration in the mid- or distal colon. et al. et al. .

2003; Topping 2003), which are retained https://doi.org/10.1079/BJN20041248 under different growing conditions (MK Morell, personal communication). In particular, the starch content of Hima- Table 6. Large-bowel digesta pH of pigs after consumption of laya 292 grain is as low as 18 %, with approximately 70 % experimental or wheat-bran (control) diets for 3 weeks (seven pigs per diet)* of the starch present as amylose compared with approxi- (Mean values and pooled standard errors of the mean) mately 50 and 25 % respectively for its Himalaya parent (Morell et al. 2003). The raised amylose content appears Colon to be due largely to a much lower content of amylopectin (Morell et al. 2003). These differences lead to substantial Diet Caecum Proximal Mid Distal changes in functional properties, including crystallinity Wheat bran 6·40a 6·61a 6·86a 6·86a and gelatinisation (Morell et al. 2003). In addition, there Namoi 6·11a 6·37a,b 6·66b 6·62b are other changes in composition, for example, an increase Himalaya 292 † 5·57b 6·09c 6·52b,c 6·57b a b,c c b in total and soluble NSP (including b-glucan), which could Oat bran 6·15 6·31 6·43 6·54 inﬂuence health-related outcomes (Topping et al. 2003). Pooled SEM 0·13 0·09 0·07 0·04 These include the lowering of plasma cholesterol and a,b,c Mean values within a column with unlike superscript letters were signiﬁ- the reduction of glycaemic index, well-established effects cantly different (P,0·05). of soluble NSP (Wursch & Pi-Sunyer, 1997; Brown et al. * For details of diets and procedures, see Table 1 and p. 608. † Mean values for six pigs. 1999; Kerckhoffs et al. 2002; Wolever, 2003). It is 612

Table 7. Large-bowel digesta total and individual short-chain fatty acid concentrations (mmol/l) of pigs after consumption of experimental or wheat-bran (control) diets for 3 weeks (seven pigs per diet)* (Mean values and pooled standard errors of the mean)

Total Acetate Propionate Butyrate

Colon Colon Colon Colon

Diet Caecum Proximal Mid Distal Caecum Proximal Mid Distal Caecum Proximal Mid Distal Caecum Proximal Mid Distal

Wheat bran 92·8a 81·6a 60·8 53·1a,b 57·4a,b 50·2 37·8 32·2a,b 26·6a 22·2a 15·4a 14·1a,b 8·8a 9·1a 7·6 6·8 Namoi 114·0a,b 107·9b 76·7 77·2a 62·1a,b 57·8 42·7 45·6a 39·9a,b 36·4b 23·7b 21·6a 12·0a,b 13·7b 10·4 10·0 Himalaya 292 128·3b 108·1b 57·9 49·2b 65·9a 56·3 33·1 30·2b 48·5b 38·8b 16·4a,b 12·6b 13·9b 13·0a,b 8·4 6·4 Oat bran 92·6a 92·1a,b 61·8 44·5b 49·2b 51·0 34·5 26·2b 31·4a 28·0a,b 17·7a,b 11·9b 11·9a,b 13·0a,b 9·6 6·5 Pooled SEM 9·6 7·2 6·5 8·4 5·3 3·4 3·4 4·8 4·5 3·7 2·6 2·8 1·7 1·3 1·1 1.2 a,b Mean values within a column with unlike superscript letters were signiﬁcantly different (P,0·05). * For details of diets and procedures, see Table 1 and p. 608. .R ide al. et Bird R. A.

Table 8. Large-bowel digesta total and individual short-chain fatty acid pools (mmol) of pigs after consumption of experimental or wheat-bran (control) diets for 3 weeks (seven pigs per diet)* (Mean values and pooled standard errors of the mean)

Total Acetate Propionate Butyrate

Colon Colon Colon Colon

Diet Caecum Proximal Mid Distal Caecum Proximal Mid Distal Caecum Proximal Mid Distal Caecum Proximal Mid Distal

Wheat bran 10·3 21·4ab 10·3a,b 5·5a 6·4 13·1 6·5a 3·3a 2·9 5·8a,b 2·6a,b 1·5a,b 1·0 2·5 1·3 0·7a,b Namoi 15·0 24·0ab 10·8a 6·4a 8·1 12·8 5·9a,b 3·8a 5·4 8·2a,b 3·5a 1·9a 1·5 3·0 1·4 0·8a Himalaya 292 15·6 27·8a 5·2b 2·6b 8·2 14·3 3·0b 1·6b 5·8 10·2a 1·5b 0·7b 1·6 3·3 0·7 0·3b Oat bran 7·7 17·9b 6·1a,b 2·1b 4·2 10·0 3·4b 1·2b 2·5 5·4b 1·8a,b 0·6b 1·0 2·5 1·0 0·3b Pooled SEM 3·2 3·3 1·8 1·0 1·7 1·6 1·0 0·5 1·2 1·5 0·6 0·3 0·3 0·5 0·3 0·1 a,b Mean values within a column with unlike superscript letters were signiﬁcantly different (P,0·05).

* For details of diets and procedures, see Table 1 and p. 608.

Downloaded from from Downloaded https://www.cambridge.org/core . IP address: address: IP . 170.106.40.139 , on on , 27 Sep 2021 at 03:13:57 at 2021 Sep 27 , subject to the Cambridge Core terms of use, available at at available use, of terms Core Cambridge the to subject , https://www.cambridge.org/core/terms . . https://doi.org/10.1079/BJN20041248 Downloaded from

https://www.cambridge.org/core

Effect of novel barley cultivar in pigs 613

Table 9. Bacterial populations (log10 colony-forming units/g) in the proximal colon contents of pigs after consumption of experimental or wheat-bran (control) diets for 3 weeks (seven pigs per diet)* (Mean values and pooled standard errors of the mean)

Diet Anaerobes Aerobes Lactobacilli Biﬁdobacteria Total coliforms Escherichia coli . IP address:

Wheat bran 9·03a 8·02a 7·93 6·84 8·00a 7·85a,b Namoi 9·14a,b 7·55b 7·39 6·27 7·47b 7·31a b a a b

Himalaya 292 9·34 8·25 8·11 6·75 8·23 8·14 170.106.40.139 Oat bran 9·24a,b 8·34a 7·49 7·31 8·22a 8·11b Pooled SEM 0·07 0·15 0·36 0·46 0·18 0·20

a,b Mean values within a column with unlike superscript letters were signiﬁcantly different (P,0·05).

* For details of diets and procedures, see Table 1 and p. 608. , on

27 Sep 2021 at 03:13:57 important to establish that these components in Himalaya digesta and faecal pH values were uniformly higher in 292 retained their physiological attributes after heat stabil- the pigs fed wheat bran than in the other three groups, isation and milling and the present experiment established which is consistent with less SCFA production in the that, in pigs, Himalaya 292 was as effective as a commer- former. Wheat bran is of relatively low fermentability, cial barley and oat bran in this regard. As expected from which helps to account for its relative superiority as a , subject to the Cambridge Core terms of use, available at human (Newman et al. 1989; Kestin et al. 1990; McIntosh faecal bulking and laxating agent (Topping & Clifton, et al. 1991), pig (Topping et al. 1993) and rodent (Jackson 2001); this would account for the greater large-bowel et al. 1994) studies, plasma total and non-HDL-cholesterol digesta mass in the pigs fed the wheat-bran diet. A study (VLDL þ intermediate-density lipoprotein þ LDL) were in human subjects has shown that the consumption of highest in our pigs fed wheat bran and lowest in pigs fed wheat bran raises faecal SCFA (Lewis & Heaton, 1997). oat-bran and Himalaya 292 diets at equivalent intakes of This occurs through diminished absorption, with no appar- NSP. Importantly, Himalaya 292 and the commercial ent change in production and explains the present faecal Namoi barley were at least as effective as oat bran. Thus, pH and SCFA data. Further, it accounts for the differences the mutation in the starch synthetic pathway does not in the proportions of the three SCFA between the pigs fed seem to have altered the effectiveness of the heat-treated wheat bran and those fed the other diets. As in previous and milled grain in lowering cholesterol. As noted, it is studies by us (Marsono et al. 1993; Topping et al. 1993) thought that this is effected largely through b-glucan. and others (Bach-Knudsen et al. 1991), large-bowel The diet in the present study was low in cholesterol so SCFA were high in the proximal large bowel and fell that the observed changes are most probably largely due towards the distal colon. This distribution is similar to to endogenous steroid metabolism rather than in choles- that seen in human surgical patients (Mitchell et al. https://www.cambridge.org/core/terms terol absorption. Although they were not measured in the 1985) and reﬂects the predominance of SCFA production present experiment, it is probable that there was an in the caecum and proximal colon and the decline of fer- increase in faecal bile acid and neutral sterol excretion in mentation (due to substrate depletion and colonic uptake) the pigs fed the barley and oat diets. This would lead to on passage of the faecal stream. Less than 10 % of the a compensatory increase in cholesterol catabolism and a SCFA produced are excreted in the faeces, which means fall in plasma concentrations (Prentice et al. 1982; Zhang that relatively small changes in transit translate to dispro- et al. 1990; Lia et al. 1995). However, it must be recog- portionate variations in their excretion (Lewis & Heaton, nised that barley contains other components that may be 1997). involved since brewer’s spent barley grain, which is Although the distributional proﬁle of SCFA within the depleted in soluble NSP, also lowers plasma cholesterol large bowel was broadly similar between groups, there .

https://doi.org/10.1079/BJN20041248 in human subjects (Zhang et al. 1991; Lupton et al. were differences by sampling site. The distribution of 1994). The relative contributions of these components to SCFA in the bowel of the pigs fed the Namoi diet was the effects of Himalaya 292 on plasma lipids remain to similar to that in those fed wheat bran, and the higher be established. Lowered serum HDL-cholesterol and values in the distal colon compared with other regions raised TAG are respectively positive risk factors for may reﬂect either differences in transit or fermentation. CHD (Coughlan & Sorrentino, 2000; Black, 2003) so the The latter appears the more probable, given the lower pH lack of unwanted effects of the Himalaya 292 diet on values. The level of inclusion of Namoi in the diet was HDL-cholesterol and TAG is potentially important if it more than double that of the other cereals and it is possible can also be demonstrated in human subjects. that there was a greater contribution of RS to large-bowel Faecal SCFA concentrations were generally similar in fermentation in this group. A more direct comparison can the four experimental groups. This is not altogether surpris- be made between Himalaya 292 and oat bran, which had ing since NSP would be a major determinant of large- similar levels of inclusion in the diet, and the higher bowel SCFA and, in the present study, the cereal products SCFA in the former support our analytical and rat data were fed in diets formulated to provide equal amounts of indicating greater RS in Himalaya 292 (Topping et al. these polysaccharides. Nevertheless, the faecal and large- 2003; Bird et al. 2004). However, the fact that the differ- bowel data suggest that there were differences between ences were conﬁned to the proximal large bowel suggests treatments in SCFA production. In particular, large-bowel greater fermentation but slower transit. Martinez-Puig Downloaded from

https://www.cambridge.org/core

614 A. R. Bird et al. et al. (2003) reported recently that a high-RS (as raw In summary, the present study has demonstrated a potato starch) diet had similar effects in pigs. Conversely, number of favourable changes of indices of cardiovascular we have reported that RS (as a high-amylose maize starch) and bowel health in pigs consuming a diet containing the appears to be fermented rapidly in pigs relative to passage novel barley cultivar Himalaya 292. Further studies are of the faecal stream (Topping et al. 1997). Indeed, it has planned to evaluate these properties in human subjects. . IP address: been shown in rats (Morita et al. 1999) and pigs (Govers et al. 1999) that for effective transport of SCFA to the distal colon, the presence of faecal bulking by NSP 170.106.40.139 seems necessary. This appears to be the case here despite References the fact that NSP were equivalent in all diets. Anderson JW, Davidson MH, Blonde L, Brown WV, Howard WJ, SCFA are known to affect intestinal morphology and Ginsberg H, Allgood LD & Weingand KW (2000) Long-term their intravenous infusion in rats reverses the colonic atro- cholesterol-lowering effects of psyllium as an adjunct to diet , on phy induced by the feeding of low-fibre diets (Topcu et al. therapy in the treatment of hypercholesterolemia. Am J Clin 27 Sep 2021 at 03:13:57 2002). Diets high in fibre or RS increase large-bowel organ Nutr 71, 1433–1438. dimensions. For RS this is either as weight (in rats; Good- Asp N-G (1992) Resistant starch. Eur J Clin Nutr 46, Suppl. 2, S1. lad & Mathers, 1990) or length (in pigs; Topping et al. Bach-Knudsen KE, Jensen BB, Andersen JO & Hansen I (1991) Gastrointestinal implications in pigs of wheat and oat fractions. 1997; Martinez-Puig et al. 2003) but it is unclear whether 2. Microbial activity in the gastrointestinal tract. Br J Nutr 65, these differences are an effect of animal species or RS 233–248. type. The apparent paradox between large-bowel hypertro- Bird AR, Brown IL & Topping DL (2000a) Starches, resistant , subject to the Cambridge Core terms of use, available at phy (which might have been expected to increase the risk starches, the gut microflora and human health. Curr Issues of neoplasms) and diminished risk of malignancy has Intest Microbiol 1, 25–37. been linked to butyrate and has been called ‘the butyrate Bird AR, Flory C, Davies DA, Usher S & Topping DL (2004) A paradox’ (Velazquez et al. 1996). It seems that this dimin- novel barley cultivar (Himalaya 292) with a specific gene ished risk reflects the effects of butyrate and, to a lesser mutation in starch synthase IIa raises large bowel starch and extent, propionate in promoting a normal cell phenotype. short chain fatty acids in rats. J Nutr 134, 831–835. The relative lack of difference in pool size of these import- Bird AR, Hayakawa T, Marsono Y, Gooden JM, Record IR, Correll RL & Topping DL (2000b) Coarse brown rice increases ant SCFA between treatments may be the reason for the fecal and large bowel short-chain fatty acids and starch but present findings where neither large-bowel length nor lowers calcium in the large bowel of pigs. J Nutr 130, mass was affected by diet. 1780–1787. The importance of the interaction between indigestible Black DM (2003) Therapeutic targets in cardiovascular disease: dietary carbohydrates and the large-bowel microflora is a case for high-density lipoprotein cholesterol. Am J Cardiol well recognised and prebiotics have been developed to 91, 40E–43E. stimulate the growth of those intestinal bacteria that are Borum ML (2001) Constipation: evaluation and management. thought to promote human health (Gibson & Roberfroid, Prim Care 28, 577–590. https://www.cambridge.org/core/terms 1995). Some forms of RS, such as high-amylose maize Brown L, Rosner B, Willett WW & Sacks FM (1999) Choles- starches, are prebiotics and stimulate large-bowel and terol-lowering effects of dietary fiber: a meta-analysis. Am J Clin Nutr 69, 30–42. faecal lactic acid bacteria but other types are not (Bird Coughlan BJ & Sorrentino MJ (2000) Does hypertriglyceridemia et al. 2000a). We did not detect any consistent differences increase risk for CAD? Growing evidence suggests it plays a in faecal and proximal colonic bifidobacteria and lactoba- role. Postgrad Med 108, 77–84. cilli and, hence, any prebiotic action of the Himalaya Gibson GR & Roberfroid MB (1995) Dietary modulation of the 292 diet in the present experiment. However, the finding human colonic microbiota: introducing the concept of prebio- of significantly lower faecal bifidobacteria numbers in the tics. J Nutr 125, 1401–1412. Namoi group before adaptation to the experimental diets Goodlad JS & Mathers JC (1990) Large bowel fermentation in rats given diets containing raw peas (Pisum sativum). Br J is a potential confounder and is unexplained. The possible .

https://doi.org/10.1079/BJN20041248 influence of differences between groups at baseline is Nutr 64, 569–587. reflected also in the borderline difference in the number Govers MJ, Gannon NJ, Dunshea FR, Gibson PR & Muir JG (1999) Wheat bran affects the site of fermentation of resistant of faecal anaerobes between the Namoi and Himalaya starch and luminal indexes related to colon cancer risk: a 292 groups during the initial adaptation period. Non-selec- study in pigs. Gut 45, 840–847. tion can be overcome by using cross-over designs rather Jackson KA, Suter DA & Topping DL (1994) Oat bran, barley than the parallel design used in the present study. Also, and malted barley lower plasma cholesterol relative to wheat the classical culture-based approach to analyse the enteric bran but differ in their effects on liver cholesterol in rats fed flora has well-recognised limitations. It is possible that diets with and without cholesterol. J Nutr 124, 1678–1684. the use of molecular tools, which have greater sensitivity Jenkins DJ, Kendall CW, Vuksan V, et al. (2002) Soluble fiber and precision, may have yielded more information on the intake at a dose approved by the US Food and Drug Adminis- prebiotic potential of Himalaya 292 and the other cereals. tration for a claim of health benefits: serum lipid risk factors for The higher numbers of faecal anaerobes in pigs fed the cardiovascular disease assessed in a randomized controlled crossover trial. Am J Clin Nutr 75, 834–839. Himalaya 292 and oat-bran diets are consistent with greater Kerckhoffs DA, Brouns F, Hornstra G & Mensink RP (2002) fermentation through substrate provision. The same is true Effects on the human serum lipoprotein profile of beta- for the difference in bacterial numbers in the proximal glucan, soy protein and isoflavones, plant sterols and stanols, colon. However, it appears that the Namoi diet inhibited garlic and tocotrienols. J Nutr 132, 2494–2505. bacterial growth, as numbers were lowest in this group. Kestin M, Moss R, Clifton PM & Nestel PJ (1990) Comparative Downloaded from

https://www.cambridge.org/core

Effect of novel barley cultivar in pigs 615

effects of three cereal brans on plasma lipids, blood pressure, Response of hepatic cholesterol, fatty acid synthesis and activi- and glucose metabolism in mildly hypercholesterolemic men. ties of related enzymes to rolled barley and oats in chickens. Am J Clin Nutr 52, 661–666. Nutr Rep Int 26, 597–604. Lewis SJ & Heaton KW (1997) Increasing butyrate concentration Roe M, Brown J, Faulks R & Livesey G (1996) Is the rat a suit-

in the distal colon by accelerating intestinal transit. Gut 41, able model for humans for studies of cereal digestion? Eur J . IP address: 245–251. Clin Nutr 50, 710–712. Lia A, Hallmans G, Sandberg AS, Sundberg B, Aman P & Spiller GA (2001) Definitions of dietary fiber. In CRC Handbook Andersson H (1995) Oat b-glucan increases bile acid excretion of Dietary Fiber in Human Nutrition, pp. 9–10 [GA Spiller, and a fiber-rich barley fraction increases cholesterol excretion editor]. Boca Raton, FL: CRC Press. 170.106.40.139 in ileostomy subjects. Am J Clin Nutr 62, 1245–1251. Stephen AM, Haddad AC & Phillips SF (1983) Passage of carbo- Lupton JR, Robinson MC & Morin JL (1994) Cholesterol-lower- hydrate into the colon. Direct measurements in humans. ing effect of barley bran flour and oil. J Am Diet Assoc 94, Gastroenterology 85, 589–595.

65–70. Topcu O, Karaday K, Kuzu MA, Ulukent S, Erkek B & Alacayir I , on

McIntosh GH, Whyte J, McArthur R & Nestel PJ (1991) Barley (2002) Enteral and intraluminal short-chain fatty acids 27 Sep 2021 at 03:13:57 and wheat foods: inﬂuence on plasma cholesterol concen- improves ischemic left colonic anastomotic healing in the rat. trations in hypercholesterolemic men. Am J Clin Nutr 53, Int J Colorectal Dis 17, 171–176. 1205–1209. Topping DL & Clifton PM (2001) Short-chain fatty acids and Marquart L, Slavin JL & Fulcher RG (2002) Whole-grain Foods human colonic function: roles of resistant starch and nonstarch in Health and Disease. St Paul, MN: American Association of polysaccharides. Physiol Rev 81, 1031–1064. Cereal Chemists Inc. Topping DL, Gooden JM, Brown IL, Biebrick DA, McGrath L,

Marsono Y, Illman RJ, Clarke JM, Trimble RP & Topping DL Trimble RP, Choct M & Illman RJ (1997) A high amylose , subject to the Cambridge Core terms of use, available at (1993) Plasma lipids and large bowel volatile fatty acids in (amylomaize) starch raises proximal large bowel starch and pigs fed on white rice, brown rice and rice bran. Br J Nutr increases colon length in pigs. J Nutr 127, 615–622. 70, 503–513. Topping DL, Illman RJ, Clarke JM, Trimble RP, Jackson KA & Martinez-Puig D, Perez JF, Castillo M, Andaluz A, Anguita M, Marsono Y (1993) Dietary fat and ﬁber alter large bowel and Morales J & Gasa J (2003) Consumption of raw potato starch portal venous volatile fatty acids and plasma cholesterol but increases colon length and fecal excretion of purine bases in not biliary steroids in pigs. J Nutr 123, 133–143. growing pigs. J Nutr 133, 134–139. Topping DL, Morell MK, King RA, Li Z, Bird AR & Noakes M Mitchell BL, Lawson MJ, Davies M, Kerr-Grant A, Roediger (2003) Resistant starch and health – Himalaya 292, a novel WEW, Illman RJ & Topping DL (1985) Volatile fatty acids barley cultivar to deliver beneﬁts to consumers. Starch 55, in the human intestine: studies in surgical patients. Nutr Res 539–545. 5, 1089–1092. Truswell AS (2002) Cereal grains and coronary heart disease. Morell MK, Kosar-Hashemi B, Cmiel M, Samuel MS, Chandler P, Eur J Clin Nutr 56, 1–14. Rahman S, Buleon A, Batey IL & Li Z (2003) Barley sex6 Velazquez OC, Lederer HM & Rombeau JL (1996) Butyrate and mutants lack starch synthase IIa activity and contain a starch the colonocyte. Implications for neoplasia. Dig Dis Sci 41, with novel properties. Plant J 34, 173–185. 727–739.

Morita T, Kasaoka S, Hase K & Kiriyama S (1999) Psyllium Wolever TM (2003) Carbohydrate and the regulation of blood https://www.cambridge.org/core/terms shifts the fermentation site of high-amylose cornstarch toward glucose and metabolism. Nutr Rev 61, S40–S48. the distal colon and increases fecal butyrate concentration in Wursch P & Pi-Sunyer FX (1997) The role of viscous soluble rats. J Nutr 129, 2081–2087. fiber in the metabolic control of diabetes. A review with special National Health and Medical Research Council (1997) Australian emphasis on cereals rich in beta-glucan. Diabetes Care 20, Code of Practice for the Care and Use of Animals for Scientific 1774–1780. Purposes. Canberra, Australia: National Health and Medical Zhang JX, Bergman F, Hallmans G, Johansson G, Lundin E, Research Council. Stenling R, Theander O & Westerlund E (1990) The influence Newman RK, Lewis SE, Newman CW, Boik RJ & Ramage RT of barley fibre on bile composition, gallstone formation, serum (1989) Hypocholesterolemic effect of barley foods on healthy cholesterol and intestinal morphology in hamsters. APMIS 98, men. Nutr Rep Int 39, 749–760. 568–574. Pachenari A, Conway PL & Playne MJ (2001) Bifidus-blood agar – Zhang JX, Lundin E, Andersson H, Bosaeus I, Dahlgren S,

. a differentiating medium for the isolation and enumeration of Hallmans G, Stenling R & Aman P (1991) Brewer’s spent https://doi.org/10.1079/BJN20041248 biﬁdobacteria from faecal samples. Biosci Microﬂora 20, 85–88. grain, serum lipids and fecal sterol excretion in human subjects Prentice N, Qureshi AA, Burger WC & Elson CE (1982) with ileostomies. J Nutr 121, 778–784.